The relationship between nuclear family structure and female achievement


Material Information

The relationship between nuclear family structure and female achievement
Physical Description:
xii, 138 leaves. : ; 28 cm.
Monson, Rela Geffen
Publication Date:


Subjects / Keywords:
Achievement motivation   ( lcsh )
Women -- Employment   ( lcsh )
Family   ( lcsh )
Sociology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Sociology -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis -- University of Florida.
Bibliography: leaves 91-95.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000577547
oclc - 13989449
notis - ADA5242
System ID:

Full Text






For Michael and Uri Zvi


I owe the attainment of this degree to many people who

encouraged me throughout the years. First, were my parents,

who encouraged their own youngest of two girls to become an

'achiever.' Second, was my husband, Michael, who for eight

years has never known a wife who was not also a student, and

who provided moral, emotional and financial support through-

out difficult years. Third, to my teachers, too numerous

to mention by name, and especially to the members of my doc-

toral committee at the University of Florida, I say thank-

you. Finally, to the chairman of my doctoral committee,

Dr. Benjamin Gorman, a wish that I may be as generous with

myself to future students as he has been to me.

Philadelphia, Pa. Rela G. Monson
November, 1972



Acknowledgements . .

List of Tables . .

Abstract . .

Chapter I Theoretical Background ..

Chapter II Methodology of the Study .

Chapter III The Demographic Background
the Respondents . .

Chapter IV Testing Hypotheses. .. ..

Chapter V Summary and Conclusions. .

References . .

Appendix . .

Biographical Sketch. . .


. ix
. 1


. 12

. 28



. o .


Table Page

III-1 Ethnic Affiliation of Fathers of Respondents 31
by Field of Respondents (in percent)

III-2 Number of Siblings in Family of Orientation 33
of Respondents by Field of Respondents (in

III-3 Religious Affiliation of Fathers by Field of 35
Respondents and for Total U.S. (in percent)

III-4 Respondents' Fathers' Educational Attainment 37
Compared with that of 35-44 Year Old Heads
of Household in 1940 and 1950 (in percent)

III-5 Fathers' Income Rank by Field of Respondents 39
(in percent)

III-6 Major Occupation Group of Head of Household q 40
for 1940 and 1950 for Ages 35-44 Compared
With Distribution of Respondents' Fathers
(in percent)

III-7 Percent of Respondents Whose Mothers Worked 42
Outside the Home After Marriage and Percent
of Respondents Whose Mothers Worked While
the Respondents Were Growing Up

III-8 Percent of Respondents Whose Mothers Worked 43
Outside the Home While They Were Growing Up
by Field of Respondent

III-9 Occupations of Respondents' Mothers Compared 45
With All Females in the Labor Force 14 Years
Old and Over Classified into Social-Economic
Groups, 1940, for the United States (in per-

Table Page

III-10 Community Involvement of Mothers by Field 47
of Daughters (in percent)

III-11 Percent of Mothers Who Worked After Children 47
by Community Involvement of Mother

III-12 Ranks of Graduate Faculties in Fields of 51
Respondents for Universities Where They
Attained the Ph.D. (in percent)

III-13 Present Academic Ranks of Respondents Con- 51
trolled for Imputed Age (in percent)

III-14 Marriage and Divorce Rates by Field of Re- 52
spondent (in percent)

III-15 Academic Achievement of Husbands by Field 54
of Wife

IV-1 Percentage Distribution of Sibling Sex Compo- 61
sition of Nuclear Family of Respondents Com-
pared with Theoretical Expected Percentages

IV-2 Sibling Sex Composition of Respondents' Fami- 62
lies and of Theoretical Population by Family
Size (in percent)

IV-3 Sibling Sex Composition of Family of Orienta- 63
tion by Field of Specialization of Respondents
(in percent)

IV-4 Percent Distribution of Family Size by Field 63
of Respondents

IV-5 Percent Distribution of Sibling Sex Composi- 65
tion of Family of Orientation by Religion of
Mother of Respondent

IV-6 Percent All-Female Sibling Families of Re- 66
spondents by Family Size by Religion of
Father, Compared with Hypothetical Distri-

Table Page

IV-7 Number of Siblings by Sibling Sex Composi- 67
tion of Respondents' Families, Controlling
for Ethnicity of Father and Compared with
Hypothetical All-Female Sibling Distribution
and Total Percent of Those Who Checked Any
Ethnic Affiliation Who Are From All-Female
Sibling Families

IV-8 Present Academic Rank for Respondents Ages 73
23-35 by Family Size (in percent)

IV-9 Present Academic Rank for Respondents Ages 73
36-45 by Family Size (in percent)

IV-10 Present Academic Rank for Respondents Ages 73
46-70 by Family Size (in percent)

IV-11 Percent Distribution of Respondents Who Are 74
Assistant Professors by Age for Only Children
Compared to Respondents With Siblings

IV-12 Percent Distribution of Respondents Who Are 74
Associate and Full Professors by Age for Only
Children Compared to Respondents With Siblings

IV-13 Percent Distribution of Number of Children in 75
Family of Orientation by Rank of School Where
Attained Ph.D.

IV-14 Percent Distribution of Respondents Who Are 78
Oldest, Middle, and Youngest Children by
Number of Siblings (Without Only Children)

IV-15 Birth Order of Respondents by Sibling Sex 78
Composition After Eliminating Only Children
(in percent)

IV-16 Is Respondent Youngest Sibling by Sibling Sex 79
Composition of Family of Orientation (Exclud-
ing Only Children) (in percent)


Table Page

IV-17 Ordinal Position of the Respondent by Rank 80
of School Where Received Ph.D. (in percent )


Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in
Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy



Rela Geffen Monson

December, 1972

Chairman: Benjamin Gorman
Major Department: Sociology

This study explores the linkages between nuclear family

structure and female achievement. Special emphasis is placed

on the dynamics of the all-female sibling family, family

size, and the meaning of ordinal position for women. It is

asserted that coming from an all-female sibling family will

be positively related to achievement, due to the resolution

by parents in favor of female achievement of the conflict

between traditional sex-role definitions and the fulfillment

of the "American Dream" through the upward mobility of their

children. A 70 percent return rate to a mailed questionnaire

yielded a sample of 485 women Ph.D.'s who held the rank of

assistant professor or above in universities granting graduate

degrees in their fields. The women came from the fields of

psychology, sociology, biology, and chemistry. Internal com-

parisons were then possible between achievers in the social

and natural sciences. In addition, census cohorts and hypo-

thetical distributions of sibling sex composition of families

of different sizes were used for comparison with the achievers.

The mailed questionnaire contained no attitudinal items, but

requested specific information concerning the structure of

the nuclear family, the socio-economic status of the family

of orientation of the respondents, maternal employment, and

a description of the family of procreation including details

of husband's education and occupation. Over 70 percent of

the women wrote comments on their family backgrounds and aca-

demic careers on the back of the instrument. These data were

coded and utilized in the analysis.

A summary of the findings follows. First, middle class

socio-economic status of the family of orientation was posi-

tively related to achievement. Second, maternal employment

was related to achievement. Diversified female role models

were also provided by mothers who were active in voluntary

community organizations. Nearly half of the mothers worked

while their daughters were growing up. Many were active

outside the home in paid jobs as well as in voluntary asso-

ciations. Third, the respondents came overwhelmingly from

small families in urban environments. There were a dispro-

portionate number of Jews. Fourth, one third reported con-

scious identification with an ethnic group. Fifth, they came

from stable homes where both parents were present while they

were growing up. All of the above general findings held for

women in both the social and natural sciences.

For the sample as a whole, there were no significant

differences in sex composition of sibling groups for female

achievers. However, when compared by ethnicity of parents

and by religious groupings, disproportionate percentages of

all-female sibling families emerged for certain groups and

family sizes. Among Jews there were a disportionate number

of all-female sibling families for three and four children

families. For affiliates with ethnic groups, those from

Western and Northern Europe came disproportionately from all-

female sibling families for families of size three, four,

and five. Though the women came disproportionately from

small families, family size was not linked to rank of school

attended for doctoral study. Eventual achievement of higher

professional rank was linked to small size of family of

orientation. With regard to ordinal position, those who

were in extreme positions were more likely to have been

achievers in the first place, but once possessing the doc-

toral degree, their attainment within their fields was un-

related to ordinal position.

The composite portrait of the life style of the aca-

demic women which emerged from the study was of women suc-

cessfully combining academic careers with marriage and

motherhood. Future research measuring ideological commit-

ment to traditional sex-role definitions by parents of fe-

male achievers and the meaning of identification with min-

ority and immigrant groups for achievement was suggested.

In addition, a parallel study of parental attitude toward

male sex-role definitions in all-male sibling families was




The sociologist chooses and delineates his area of study

within the boundaries of societal conceptions of his subject

matter area. American society is highly concerned with its

open-class qualities which lead to the dream of upward mobil-

ity for all. Study of the structural and cultural factors

which lead to achievement of particular groups within the

society, or the lack of achievement by others, has been a

focus of research. However, these studies have most often

concentrated on achievement of men and have ignored women.

Perhaps the paucity of meaningful data on female achievement

is due to sex-role definitions which persist in our society

and which, in effect, define achievement as an aspect of

the male role. This theory of sex-role is best expressed

in the Parsonian (Parsons,1955:23) formulation of the di-

vision of labor by sex within the family, with the woman

performing the expressive role while the man is the instru-

mental leader. This study is specifically concerned with

the linkages between nuclear family structure and female

- 1 -

- 2 -

achievement. Special emphasis will be placed on the dynam-

ics of the all-female sibling family, as well as the mean-

ing of birth order for women.

We will begin with the hypothesis that parents have a

great effect on the socialization of their children, includ-

ing children's achievement orientation and motivation. In

a recent survey of the sociological literature on the family

in the last decade (Walters and Stinnett,1971:120) it is

noted that

Consensus existed among the studies reviewed that
academic achievement, leadership, and creative
thinking of children was positively related to warm,
accepting, understanding, and autonomy-granting
parent-child relationships. The studies also indi-
cate that parents of children who are high academic
achievers tend to value and encourage academic

In his discussion of family structure and educational at-

tainment, Elder (1969:51) puts it succinctly:

Family structure is one of the more important de-
terminants of achievement motivation and skills.
Many of the personal qualities and skills that en-
able children to meet standards of excellence ...
are acquired in parent-child relations providing
guidance and yet allowing the child freedom to
develop independent mastery and responsible deci-

We assert that parents have certain role expectations

for sons and daughters and that these expectations usually

differ. This is so despite the fact that boys and girls

- 3 -

are often exposed to similar educational situations. Child-

ren's perceptions of the desirability of certain sex-linked

activities have been studied even in five year olds (Fauls

and Smith,1956:105-17). In that study, five year olds of

both sexes were shown a series of paired pictures showing

a sex-appropriate and a sex-inappropriate activity. Both

their personal preferences and their beliefs about the acti-

vity that mother and father would prefer for boys and girls

were investigated. Boys chose 'masculine' activities more

often than girls did, and children of both sexes indicated

that the parents preferred the activities appropriate to the

child's sex more often than the sex-inappropriate activities.

This issue of sex-role learning is discussed in a broad way

by Williams (1970:80) where he states that

It is a commonplace observation that in our culture
there are many similarities in the values parents
attempt to develop in boys and girls, for example,
independence and social maturity. But there are im-
portant differences also. Girls are more likely
than boys to emphasize being well liked; to value
interpersonal harmony and success; to stress the
"tender virtues" of kindness, consideration, sympathy,
and understanding; to attach great importance to
moral values and aesthetic considerations.

Our society is a success oriented one. Achievement mo-

tivation is considered crucial to maturity, and parents

dream and scheme for the upward mobility of their offspring.

- 4 -

As Williams (1970:454-5) notes,

All societies have standards of character and pro-
ficiency, and accord rewards to those best meeting
whatever standards are most highly appraised, whe-
ther of military prowess, ritual knowledge, asceti-
cism, piety, or whatnot. The comparatively strik-
ing feature of American culture is its tendency to
identify standards of personal excellence with com-
petitive occupational achievement.

Indeed, parents who themselves are successful in a limited

way or who have personally defined their own experience as

failure "may mute their original goal-emphasis and may defer

further efforts to reach the goal, attempting to reach it

vicariously through their children" (Merton,1961:159). In

accord with this desire for achievement and the differential

expectations for boys and girls, parents usually project

this expectation upon their sons rather than their daughters.

This fact is documented well by the psychological and socio-

logical literature on achievement which deals almost exclu-

sively with the origins of achievement orientation in males.

The acknowledged classics in the field, for example, the

study by Rosen and D'Andrade (1969) of the "Psychosocial

Origins of Achievement Motivation," concern males only. In

a discussion of this same point, Matina Horner, a psycholo-

gist, (1970:50) states that:

Even more striking is the absence of any mention of
achievement motivation by McClelland (1961). Using

- 5 -

evidence from vases, flags, doodles, and children's
books, he was able to study achievement motivation
in such diverse samples as Indians, Quakers, and
Ancient Greeks but not in women. This was not an
oversight ... there are in fact not many meaningful

We have already speculated that the lack of meaningful data

is due to the linkage of the ideal of achievement with the

male role. We argue that despite this sex-role stereotype,

the parents' desire for achievement on the part of their

children is so strong that in the absence of a male child

the parents will allow, encourage, or push daughters into

the achievement role. Psathas (1968:260-1) suggests that

The rationale for the present consideration of sib-
lings together with family finances lies in the
view that education and training can be expensive,
that work is generally defined as being the primary
responsibility of males and that consequently, be-
cause education is seen as a major determinant of
social mobility, the education of sons for occupa-
tions is expected to predominate over that for
daughters. Thus, with given levels of income, the
chances of a girl entering an occupation that in-
volves expensive and lengthy preparation will de-
pend on the presence or absence of male siblings,
their proximity in age and their birth order in re-
lation to the girl.

In summary, one of our hypotheses is that coming from an all-

female sibling family will be positively related to achieve-

ment. Within this framework, birth order and family size

may determine which sister achieves. This hypothesis sug-

gests that it is the structural variables rather than per-

- 6 -

sonality traits which lead to achievement motivation in fe-

males, and affirms the primacy of sociological over psycho-

logical determination.

In an article which summarizes the decade's research

(Walters and Stinnett,1971:123-4), the importance of ordinal

position is noted.

The literature indicates that parental responses to
children are a function of the ordinal position of
the child and size of the family. The findings sug-
gest that parents tend to be more supportive of and
also tend to exert more pressure for achievement
upon first-born children.

Since most of the work on birth order concerns males only,

the effect of ordinal position on females is unclear. Some

theorists (Epstein,1970:79; Kammayer,1966) suggest that be-

ing first-born works against the autonomy of the girl.

Parents might be more apt to chart a known course
for their first-born; and for women in most cases
this would be a traditionally defined sex role ...
the older daughter might easily become a surrogate
mother in her family. (Epstein,1970:79)

Kammayer (1966:508-16) found that first-born girls are more

traditionally oriented toward the female role. However, he

never specified whether his first-born girls who were more

oriented toward traditional roles came from mixed-sex sib-

lings or all female families. In a study of birth order

and responsibility (Harris and Howard,1968) the most consis-

- 7 -

tent finding was that the first of sex--in comparison with

the later of sex--endorsed an earlier age for assumption of

responsibilities on the part of children. Also, "As a

group, the girls gave earlier responsibility and indepen-

dence scores than did the boys" (Harris and Howard,1968:

431). This factor operated independently of birth order.

The authors then go on to attribute seriousness of purpose

and career commitment for first of sex males to high respon-

sibility orientation. They ignore the anomalous finding for

females of higher responsibility scores and lower career

commitment. The above data suggest that birth order may

have a different meaning for males than for females and

within the context of an all-female versus a mixed-sex sib-

ling structure.

Ordinal position may be linked to achievement for fe-

males, but as implied above the advantages of being the

eldest are unclear. In fact, it may be that the youngest

in all-female sibling families is the most likely to

achieve. Her definition by the parents as the last child

entails the recognition that they must fulfill their

achievement desires through a daughter. The elder sisters

may have already been socialized to a more traditional fe-

male role, leaving the youngest the most latitude as to

- 8 -

role definitions. We hypothesize some patterned relation-

ship between ordinal position and achievement within the

all-female structural context. However, the exact nature

of this relationship and its similarities and divergences

from that of men's birth order and achievement is left to

emerge from the data.

Two other major relationships are to be investigated.

The first concerns the relationship between family of orien-

tation's socio-economic background and female achievement.

Secondly, we anticipate a positive relationship between ma-

ternal employment and female achievement. We expect that

most achievers come from middle-class backgrounds. Indices

of this background which are commonly used in the literature

are the education of the parents, the income level of the

family, and the occupation of the father. Higher levels of

education, income, and occupational rank are positively re-

lated to achievement (Turner,1962; Psathas,1968; Sewell and

Shah,1968). Walters and Stinnett (1971:24) state that

The literature indicates that basic differences
exist in parent-child relationships according to
social class which reflect different living con-
ditions. The studies generally indicate that
middle-class parents tend to be both more suppor-
tive and controlling of their children, and that
they are more likely to discipline their children
by utilizing reason and appeals to guilt and are
less likely to use physical punishment than are
lower-class parents.

- 9 -

It is clear that social access to the means for achievement

in American society is unequally distributed. Wealth, race,

social status, and sex all have an effect on opportunities.

Values vary in different segments of society as well, values

that are passed on from parents to children. Coser (1969:

xi, xiii), in her introduction to Life Cycle and Achievement

in America, notes that

Insistence on character building and personality
growth governs the outlook of the middle and upper-
middle class parents, who prepare their children
for occupations in the professions and executive
work ... Child-rearing practices that are focused
on the use of external authority not only are con-
gruent with the occupational roles available to
the members of the working class but also help to
de-emphasize the notion of individual achievement.

We expect to find a relationship between middle-class status

of the nuclear family and female achievement. Female

achievers' backgrounds should include parents with higher

educational background than their peers, fathers with higher

income than their peers, and father and mothers whose occu-

pations were in the upper status ranks of the American stra-

tification system.

Maternal employment has often been viewed in the past

as detrimental to the stability of the home, and to chil-

dren's development. However, a recent compilation of 22

empirical studies on the employed mother (Nye and Hoffman,

- 10 -

1963) has shown that maternal employment has no unfavorable

effects on children. The important factors are the mother's

reasons for working, the quality of the care which the chil-

dren receive in her absence, and the attitudes of her hus-

band. Our concern is with the positive relationship between

maternal employment and female achievement. In his study

of Life.Styles of Educated Women, Ginzberg (1966:29-30) ex-

plored the employment backgrounds of the mothers of his re-

spondents and the effects of such employment on them. He

found that the greatest influence was reported by women whose

mothers had worked after they were married.

In fact 2 out of every 3 girls whose mothers had
worked after marriage reported that this had an
effect on their own plans. Almost all saw the in-
fluence as positive. While they were growing up
they looked forward to emulating their mothers
and combining home and work.

Our study explores the variables of maternal employment after

marriage and maternal employment while raising children.

Several other studies (Psathas,1968; Astin,1969; Almquist

and Angrist,1970; Harmon,1972) have found positive relation-

ships between having a working mother and the subsequent

decision by their daughters to include work and homemaking

in their own lifestyles. We also explore the community ac-

tivism of the mothers of our respondents. Our supposition

is that voluntary community involved -nt could provide a

- 11 -

model of a broader female role than homemaker for daughters,

just as the outside work of a mother for remuneration does.

Finally, we speculate that achieving women who combine the

roles of domesticity and career may be married to men whose

mothers have worked. Such men, we feel, would tend to be

more supportive of their wives careers than the sons of women

who did not work after marriage.

We have introduced the major foci of our study: nuclear

family structure of the achieving female, socio-economic

background of her family, type of role model provided her

by her mother, and type of spouse role-expectations provided

her husband by his mother. We shall now turn to the metho-

dological approach used to collect the data needed for test-

ing our hypotheses.


Our study of female achievement necessitated the col-

lection of specific information about the nuclear family

structure of women who were achievers. In addition, we

needed some data on their families of procreation. Inher-

ent in the previous statements was the need to define the

term 'achievers' so that the parameters of our population

would be clear, and an appropriate sampling frame could be

devised. The population defined was one which consisted of

all women fulfilling certain criteria within several fields

of academia. The academic community was chosen for study

because it internally specifies several criteria of achieve-

ment. The first of these is the doctoral degree, and the

second is the rank-title scale used for professors. Our

sample was restricted to those holding doctorates and current

employment in an academic institution which had a graduate

program in the specialty of the women to be sampled. The

percent of women who had met these standards of achievement

in the academic world was small. Thus, the ones who had

- 12 -

- 13 -

'made it' in the academic world could be assumed to have

exceptional persistence and ambition and could clearly be

called achievers. Women were to be chosen from four sub-

fields within the university: two social sciences and two

natural sciences. This decision was based on the notion

that women who chose careers in fields considered 'most

masculine' by society as a whole would be an 'ideal type'

of achieving females and might differ systematically from

those who achieved in fields more socially acceptable for

women. The specific natural and social sciences were chosen

partially because of the availability of directories of

their members which facilitated contact with them through

the mails. The women chosen from the four fields were

viewed as members of four sub-samples of the population of

women who had achieved through academic careers.

We decided to select a national sample in the interest

of broad applicability of results. In order to reach a

large number of women with limited financial resources at

our disposal, we utilized a mailed questionnaire which was

simple and brief. The selection of women from the four sub-

fields of psychology, sociology, biology, and chemistry in-

sured that some meaningful comparison groups would be built

into the sample, though a control group of non-achievers

was not to be sampled.

- 14 -

Sampling Procedure

The psychology sample was drawn from the 1970 biograph-

ical directory published by the American Psychological Asso-

ciation. Graduate departments were marked off in the section

which divides the membership by schools. Schools which

granted graduate degrees were ascertained through Graduate

Study in Psychology 1971-72,which is also published by the

American Psychological Association. Pages to be used were

selected by three-digit random numbers taken from a table

of random numbers. These pages were from the geographical

and institutional directory of membership found in the back

of the larger directory. All schools on the page so chosen

had all of their female Ph.D.'s with a rank of assistant

professor or above selected for the sample. Each name se-

lected was checked in the biographical section of the direc-

tory in order to eliminate all those below the rank of as-

sistant professor, with foreign degrees, or with degrees in

education. We continued sampling pages until 225 were se-

lected. Of these, 27 had doctorates in education and were

eliminated from the sample leaving 198. The problem of

names will be discussed later for all the groups together.

The decision to eliminate those with foreign degrees from

consideration was based on the assumption that socialization

- 15 -

patterns differ in different cultures, and that those with

degrees from abroad were likely to have been raised with

different sex-role definitions and expectations than our

American sample. Thus, leaving them in might bias the

sample. Those who were raised in foreign countries and who

went to graduate school in the United States were eliminated

later when returned questionnaires revealed that they at-

tended high school abroad. (See upcoming section on return


The sociology sample was drawn from Guide to Graduate

Departments of Sociology 1971, which is a publication of the

American Sociological Association. The sample consisted of

all women listed in the guide who had their doctorates in

sociology and the necessary rank of assistant professor or

above. Thus, women with joint appointments with other de-

partments were selected for the sample if a check with the

larger American Sociological Association membership directory

showed them to be members of the association and to have com-

pleted their graduate degrees in sociology. A total of 206

women were selected in this manner.

The chemistry sample was originally drawn from the

International Chemistry Directory 1969-70. It included all

women listed in the directory who met the aforementioned

- 16 -

criteria of achievement in chemistry. This list was then

cross-checked with the American Chemical Society Directory

of Graduate Research, 1969, and the 147 women in schools

with graduate programs were selected out of the original

list. The women teaching in colleges of osteopathy were

included in the sample although they did not appear in the

graduate directory. Because of the shortage of women in

graduate departments of chemistry, those in chemistry and

biochemistry were included as long as.they were listed in

the American Chemical Society directory.

The biology sample was drawn from the Annual Guides

to Graduate Study, 1971, Book II: Biological and Health

Related Sciences. Women were selected from the following

sections: virology, botany, cellular and molecular biology,

microbiology, genetics programs, pathology, physiology, and

zoology. They were chosen through random selection of pages

using a three-digit random number from a table of random

numbers. All women on a page (one school) were then se-

lected for the sample. In this manner 212 women were cho-


In the selection procedure in all fields certain first

names were automatically eliminated because they are given

to both men and women. Thus, people with the names of Lynn,

- 17 -

Jean, Francis, and Marion were not selected for the sample.

(Despite this precaution, we received notes from 13 males

who were kind enough to specify their sex and to let us know

that they should be eliminated from the sample.)

The completed list of 762 respondents was then com-

piled. Each was given an identification number in which

the first digit signified field of expertise. Question-

naires were mailed to them at their university addresses.

It was felt that respondents were more apt to answer mail

received at their offices. Also, university addresses were

more permanent than home addresses and secretarial help

would lead to forwarding of questionnaires to those on leave

or at new posts.

The Final Sample--Return Rate

The questionnaires were mailed out on March 1, 1972.

In all, 762 were mailed with the following breakdown: 198

psychologists, 206 sociologists, 146 chemists and 212 bio-

logists. The mailing was timed so that there would be ade-

quate time for response before spring vacations. The re-

sponse was swift and gratifying. Of the questionnaires later

judged usable for the study, 457 or 94 percent had been re-

turned by March 30, 1972. The response rate was computed

- 18 -

for each of the fields and for the total sample. Several

factors were taken into account. Questionnaires which were

returned with notes stating that the respondents were male

were eliminated from the sample size. Those who specified

that they had attended high school abroad were also elimin-

ated. Three were returned with notes stating that the re-

spondents were deceased. These were also eliminated as were

retirees and professors-emeritus who wrote back and stated

that they did not wish to fill out the questionnaire. Those

who sent letters of refusal or whose instruments were re-

turned by the post office as wrong addresses were tabulated

but kept in the sample size. Here are the results of this

tabulation by field.

Of the 198 psychologists who were mailed questionnaires,

124 returned usable ones. In addition there were 13 wrong

addresses, 3 foreigners, 1 male, 2 retirees, and one who re-

fused to fill out the instrument. Thus, a total of 144 re-

spondents were accounted for. The final sample size after

eliminating foreigners, males and retirees was 192. This

yields a 65 percent return rate for the psychology sub-


Of the 206 sociologists who were mailed the question-

naire, 138 returned usable ones. In addition there were 2

- 19 -

wrong addresses, 17 foreigners, 5 males, 3 refusals, and

one received after the coding deadline (June 1, 1972).

Thus, a total of 166 respondents were accounted for. The

final sample size after eliminating foreigners, males, and

retirees was 184. This yields a 74 percent return rate

for the sociology sub-sample.

Of the 146 chemists who were mailed the questionnaires,

80 returned usable ones. In addition there were 12 wrong

addresses, 7 foreigners, 7 males, 2 deceased, and one mailed

in after the coding deadline. Thus, a total of 109 respon-

dents were accounted for. The final sample size after elim-

inating foreigners, males and deceased was 130. This yields

a 61 percent return rate for the chemistry sub-sample.

Of the 212 biologists who were mailed the question-

naire, 143 returned usable ones. In addition, there were

2 wrong addresses, 14 foreigners, 2 males, 1 deceased, 1

retired, and 1 received after the coding deadline. Thus,

a total of 164 respondents was accounted for. The final

sample size after eliminating foreigners, males, retirees,

and deceased was 194. This yields a 73 percent return rate

for the biology sub-sample.

The final sample size after eliminating males, for-

eigners, retirees, and deceased was 700. Of this total,

- 20 -

485 usable questionnaires were coded and appear in the

tables. This yields a total return rate of 69.4 percent.

Had we extrapolated the eliminated categories to those who

didn't return the questionnaire according to the same per-

centages in which they appeared in the returned sample, the

return rate would have been well over 70 percent. The so-

ciologists and biologists had the highest return rate, seem-

ingly eliminating any explanation of variation in rates by

natural versus social sciences. We have no reason to sus-

pect that our respondents differ systematically from the

non-respondents as our independent variables relate to nu-

clear family structure and not to attitude. The later ar-

riving questionnaires did not differ from those received

immediately. In any case, as previously noted, 94 percent

were received within the first month. Those questionnaires

which were returned were fully filled out and were often

accompanied by notes, letters, and even articles for the re-

searcher. In addition, 79 percent of the respondents made

comments on the back of the questionnaire in the space pro-


Major Variables

Through the questionnaire, we sought to obtain suffi-

cient data to test our hypotheses and to compare our sample

- 21 -

with those tested by others in studies of birth order and

of achievement. In some cases, we wanted more complete in-

formation than had previously been gathered in such re-

search. Thus, we needed complete information on the struc-

ture of the family of orientation including size of family,

spacing and sex of children, birth order of children, place-

ment of respondent, and educational and occupational achieve-

ment of siblings. With regard to respondents' parents, we

requested information on religion, ethnicity, family stabil-

ity, educational attainment, work records of mother and

father, community activism of the mother, and a rough income

ranking of the family. For each respondent we asked academic

rank, year of completion of the doctorate, and the institu-

tion at which the doctorate was completed. In addition, we

asked whether the respondent attended high school in a rural

or urban area, and in which section of the country she grew

up. With regard to family of procreation we asked the re-

spondent's marital status, and the educational and occupa-

tional status of her husband. We asked if the husband's

mother ever worked. Finally, room was left for comments, and,

as noted above, 79 percent of the women utilized this option.

No attitudinal questions were included in the instru-

ment, although some attitudinal data were collected through

- 22 -

the comments. This commission was a conscious decision based

on sociological and financial considerations. We were con-

cerned with structural antecedents of female achievement,

not-perceived components of success. It was believed that

the objective questions would suffice for analysis of our


Creating A Control Group

Several of the major hypotheses of the study required

a control group for verification. For example, we could not

state that we had a "disproportionate" number of achievers

from all-female families if we had no idea of what the actual

distribution of family sizes and sex make-up was in the larger

population of the United States. Similarly, we needed to

know income distributions of the United States, number of

women married in the total population of our respondents'

age cohorts, and the fertility of this same group. The de-

mands of the above hypotheses led to the use of census data

to create a hypothetical control group for the study. As

previously noted, for the researcher to undertake a survey

of a control group of non-achievers would have involved pro-

hibitive expense. Although comparisons could be made among

the women in our four sub-fields, we sought a larger and

more meaningful population for major comparisons. At issue

- 23 -

were the special or distinctive qualities of achievers' nu-

clear families, and the ways in which achievers' families

differed from those of non-achievers, or the general popu-

lation. Data on family properties of the general population

were available from the United States Census.

Different census years were utilized for comparisons

with figures for our respondents. Questions which related

to the family of orientation of the respondents required the

use of comparative figures which approximated the distribu-

tion of the United States population when respondents were

growing up in their families of orientation. Since 80 per-

cent of our respondents were between the ages of 25 and 45,

fertility data for their mothers' cohorts were taken from

the 1950 census, approximating the decade in which most of

the mothers of our respondents completed their child-bearing

years. Similarly, data on occupations and educational

achievement for comparison with their mothers' and fathers'

educations and occupations were taken from 1940 and 1950

census material for men and women of middle age. For compar-

isons involving our respondents own characteristics and those

of their families of procreation, the 1960 and if possible

the 1970 census was used.

In the discussion of the hypotheses relating to all-

female sibling nuclear families of orientation, a hypothetical

- 24 -

statistical distribution of such families in the population

was computed mathematically. This distribution took into

account the fact that one member of each of the families in

our sample was fixed as female. Ideal percentages of all-

female versus mixed-sex sibling families were computed for

each family size and used as a base line for comparison with

the families of orientation of our respondents.


Since there were no attitudinal data included in the

instrument, most of the coding was quite straightforward.

However, several problems did have to be solved. All ques-

tions about occupation were coded using the 1970 census oc-

cupation code. Several additional coding categories, such

as student and housewife, were added to this code so that

we could usefully categorize all of the major occupations

of our respondents and their families. The census occupation

code was then collapsed into thirteen categories: profes-

sionals, executives, semi-professionals, managers and pro-

prietors, technical workers, sales personnel, clerical

workers, craftsmen, operatives, laborers, service workers,

students, and housewives. The census region code was used

for categorizing regional distribution of respondents.

- 25 -

City size was calculated through the use of a 1950 atlas and

then coded by size categories to yield an approximate break-

down of respondents by rural and urban residence while they

were growing up. School rankings for universities where

our respondents received their doctoral degrees were computed

through the use of Roose and Anderson's (1970) ratings of

graduate programs. When women were sampled who had completed

doctorates in sub-fields of a discipline such as chemistry

or biochemistry, composite rankings were computed. Inadver-

tantly, the age of the respondent was not included in the

personal data in the instrument. An approximate measure of

age was computed from the year that they completed the Ph.D.

This approximate measure was then used as the control vari-

able for such tables as comparisons of professorial rank at-

tained by women with different backgrounds.

Not only did many women take advantage of the option

of making comments on the back of the questionnaire, but

they often added sheets of comments and letters to the in-

vestigator. This outpouring of data about their own personal

decisions to pursue academic careers and to obtain the Ph.D.

degrees was an added source of information for us. There-

fore, a code was developed for these qualitative data. The

code emerged from the comments themselves rather than being

- 26 -

superimposed upon them. Thus, any person or situation men-

tioned repeatedly by the women as significant in their aca-

demic development was coded. The actual codes developed

and the marginals for the entire sample will be found in

the Appendix.

Statistical Analysis

Most of the data collected were nominal although some

such as professorial rankings, occupational classifications,

and graduate faculty ratings of universities were ordinal.

The key distinctions used in the analysis, such as achiever

and non-achiever, were nominal in character. They were

tested through the use of X2 as a measure of goodness-of-fit

when proportions of achievers from all-female sibling fami-

lies were compared with a hypothetical distribution. X2 as

a measure of independence was used when testing the nature

of the relationship between such variables as ethnic affil-

iation and family size.

In the case of the data regarding sibling group sex

composition, chance probability was calculated by assuming

that the probability of the birth of a daughter as against

that of a son was 50 percent. Thus, the probability that

a sibling group of n children would be composed solely of

- 27 -

females, including the respondent whose sex was known, was

p = (n-l)

We have now described the theoretical background of the

study together with some of the hypotheses which we sought

to test. In addition, in this chapter the methodological

execution of the study was traced. Let us now turn to the

results. We shall begin by describing the women who responded

to our questionnaire and thus enriched our knowledge of aca-

demic women.


While this chapter is devoted mainly to the demographic

parameters of the sample group, these descriptions become

meaningful only against a backdrop of the parameters of the

general population from which they came. In some cases this

sample comparison with the larger population is also a test

of hypotheses concerning the background of academic women.

This is particularly evident in the discussion of the re-

spondents' socio-economic background and in the discussion

of their mothers as role models.

Women chosen on so stringent criteria of achievement

as that which we employed clearly would differ from their

age cohorts in the general population. But how? Would they

be products of broken or stable homes? Would they combine

marriage with career achievement and motherhood as well?

In this chapter we present the first and most general pic-

ture of the parameters of this unusual group of women.

Our respondents ranged in age from their mid-twenties

to early seventies: 80 percent were between the ages of

- 28 -

- 29 -

25 and 45. Age distribution did not differ significantly

by sub-field. They came, overwhelmingly, from urban back-

grounds. In fact, 32 percent grew up in cities with over

a million in population and only 15 percent grew up in towns

with under 5,000 inhabitants. These figures stand in con-

trast to the general population in the 1940 and 1950 cen-

suses which serve as our imputed base cohorts. In 1940,

56.3 percent of the national population was classed as urban

and in 1950 64 percent of the national population was so

classified. According to Smith (1960:97), this was partially

due to a change in definition. If the same definition had

been used for both years,then the 1950 national population

would have been classified as 59 percent urban. Using com-

parable definitions, at least 85 percent of our respondents

grew up in urban settings. We can focus this more starkly

by noting that 32 percent of our respondents grew up in

cities with 1,000,000 or more inhabitants, while, according

to the 1950 census (U.S. Bureau of the Census,1952:6,7),

only 11.5 percent of the total national population then lived

in cities of over 1,000,000. In comparing the sample's dis-

tribution by regions of the country with that for the total

United States population in 1950, we found that the Northeast

and Middle Atlantic states were over-represented in the

- 30 -

sample. This was probably due to the fact that certain re-

gions of the country were more urbanized than others and our

respondents come disproportionately from urban settings (Dun-

can and Reiss,1956:29).

Comparing the several sub-fields, women from all of the

sub-fields came from urban backgrounds. The psychologists

were the least likely of all to come from rural areas, with

only 8 percent from towns of under 5,000 population. With

regard to regional distribution, there was no divergence

from the general trend toward the Middle Atlantic and East

North Central states by sub-field, except for the exception-

ally high percentages of women biologists and chemists from

the Middle Atlantic states (43 percent and 41 percent respec-

tively). Complete marginals for the total sample on age dis-

tribution and rural-urban and regional distribution will be

found in the Appendix.

Since over 98 percent of our respondents were Caucasian

we shall consider them a "white" sample for census compari-

sons. One-third of the women identified an ethnic group

with which their parents considered themselves affiliated.

Among alternatives offered, and frequently checked, were

Italian, Irish, Polish, and German. In addition, some re-

spondents checked "other" and many wrote in ethnic groups

- 31 -

not on our offered list. Two such groups were volunteered

so often that they were coded separately: English and Jew-

ish. We had included Jewish as a category under religious

affiliation, but did not offer it as an option under the

item on ethnic affiliation. Despite this, 25 percent of

all those who reported an ethnic identification wrote in the

category of Jewish and checked it. The "other" category

which was not recorded consisted mainly of Northern and West-

ern Europeans groups together with a small number of Slovaks,

Australians, Lebanese and Canadians. Table III-1 gives the

distribution of ethnicity by sub-field for the fathers of

the respondents, after excluding the 69 percent of the sample

who checked no ethnic identity for father.

(in percent )

GROUP Psychology Sociology Chemistry Biology

Italian 6 2 4 10
Irish 14 14 9 5
Polish 9 2 13 10
German 20 24 26 20
Other 26 22 35 23
Jewish 23 37 13 20
English 3 0 0 13
(N) (35) (51) (23) (40)

- 32 -

More of the sociologists came from families with ethnic af-

filiations than any other group. This difference was largely

accounted for by those sociologists who wrote in the category

Jewish. The marginals for the total sample for ethnic af-

filiation of mothers and fathers will be found in the Appen-


Most of our respondents grew up in homes with smaller

families than was the norm for their cohorts in the 1940

census. Two-child families were the modal nuclear family

structure in the sample (36 percent) while in the general

population two-child families accounted for only 22 percent

of the total. As with regional distribution, this may be

related to the highly urban concentration in the sample.

According to T. Lynn Smith (1960:311),

Among the white population in every one of the
forty-eight states, for the period between April,
1945, and 1950, the fertility ratio of the urban
population is substantially lower than those for
the rural-nonfarm and the rural-farm categories;
... The tendency of the birth rate of the rural
population of the United States to exceed that
of the urban has been known, of course, for some
time, ... Nevertheless, it is important to see
the degree to which it persists throughout the
length and breadth of the country, even after
the upsurge in the urban birth rate since 1935
and after the elimination of so much of the dif-
ference between the rural and urban ways of life
that has accompanied the perfection of modern
means of communication and transportation.

- 33 -

Still, more than half of our respondents grew up in homes

with three or more children. Table III-2 shows the break-

down of family size by sub-field of respondent. With the

exception of the fact that more of the biologists than any

other group come from large families, field of specializa-

tion is unrelated to family size.

(in percent )

1 2 3 4 5-8 (N)
Psychology 21 39 21 10 10 (124)
Sociology 17 38 20 14 10 (138)
Biology 18 32 20 13 17 (143)
Chemistry 14 36 26 15 9 ( 80)

Approximately one-quarter of the respondents reported that

their family was separated for some length of time while

they were young. In most of these cases the father was ab-

sent. The most common reason for such separations for all

sub-fields was the death of one parent (39 percent). Next

in import was divorce,which accounted for 23 percent of the

broken homes. Other separations were a product of illness

of a parent, or separation from the father because of war

or army services on a career basis.

- 34 -

Close to 60 percent of the respondents reported that

their parents were affiliated with a Protestant denomination.

Another 12 percent of the fathers were Catholic, 22 percent

were Jewish, 2 percent other religions, and 7 percent wrote

in that their fathers had no religion or were agnostic.

Several of those who wrote in "no religion" or "agnostic"

had also written in Jewish under ethnicity. Table III-3

shows the religious affiliation of the fathers by the field

of the respondents. It also shows the approximate religious

breakdown of the United States population according to a

1957 survey conducted by the Bureau of the Census. It is

clear that Jews are disproportionately represented in this

sample, though the degree of disproportion varies. The so-

cial sciences have more Jews than the natural sciences. It

has been estimately that there are between five and six mil-

lion Jews in the population, constituting less than 3 percent

of the total population of the United States (Fine and Him-

melfarb,1971:5). The large percentage of Jews in the sample

may contribute to the small family size and large urban re-

sidence of the sample. Jews are generally supposed to have

the lowest birth rate of any group in America, and to be the

most concentrated in urban areas. (See Smith,1960:330; Gold-

stein,1971:15,34,38.) According to the census figures of

- 35 -


m O
% <


I 0 r
H o
H 0H H

H z

4 0

H r


0 -
rO 0

0 r.


r- r- 11 o

M r-I o

In un co
N N r-- r-I

m CN l H o

r-i t It D rA
in in us Ln

>1 Q04
w >1 >1 024n
o a,
0 t0) 4 <
-I 0 >i 4J 0
O r- to ) *.dq
S 0 0 *r- 34 J
u -H r-H n rd
Ui 0u I c: 0 0)
(0 0 U H 0 *H
4 U : CT U

- 36 -

the 1957 survey, approximately 26 percent of the population

of the United States was Roman Catholic, while 66 percent

were members of various Protestant denominations. According

to the above figures, the Roman Catholics were most under

represented among our respondents. The small family size

for the group as a whole reflected this, also, since Roman

Catholics have the highest fertility rate of the three major

religious groups in the United States.

Socio-Economic Background

An important background characteristic of our respond-

ents is the socio-economic status of the homes from which

they came. Several indicators of this status were included

in the questionnaire. The first concerned highest level of

educational attainment of both the mother and the father of

the respondent. (See Appendix for marginals for total

sample.) We found that over one-third of all the fathers

had completed at least a bachelor's degree, and 22 percent

of all the fathers possessed a graduate degree. One-quarter

of the mothers had at least a bachelor's degree. At the

other extreme, 31 percent of the mothers and 32 percent of

the fathers had not completed a high school diploma. If we

compare the figures for our respondents' families with those

for the general population, we find that in 1940 and in 1950

- 37 -

the educational level of the heads of household was much

lower than that for the fathers of our respondents. Table

III-4 makes this comparison. The census data are from

American Families (Glick,1957:89).

(in percent )

EDUCATIONAL Fathers of Fathers of Fathers of
ATTAINMENT Respondents 1940 Cohort 1950 Cohort

School 18 59 41

Some High
School 14 17 20

High School 14 12 21

Some College 18 6 8

Four Years or
More of College 35 6 9

No Answer 2 1 2

Our respondents' parents, from all fields, are more highly

educated than the general parental generation.

A second indicator of socio-economic status of the fam-

ily is the income rank of the fathers. Here the variant

ages of the respondents made it difficult to ask an objective

- 38 -

question on salary. The meaning of a $5,000 a year income

in 1930 differed substantially from its meaning in 1950--

and we had respondents who would report their father's in-

comes from time periods varying from 1920-1960. Also, re-

collection of exact parental income or realistic estimation

of it by the respondents was unlikely. Therefore, we de-

vised a question which would be relative to a well-known

status in the economy to measure income. We asked: "When

you graduated from high school would you say that your

father's income was: less than that of a high school teacher

at that time, about the same as a high school teacher at

that time, or more than that of a high school teacher at

that time." The position of high school teacher was taken

as a rough indicator of middle class status and middle in-

come. Of the 459 respondents who answered the question, 21

percent reported their father's income as less, 18 percent

as the same and 49 percent as more than a high school teacher

at that time. This scatter of respondents suggested that

our indicator question was sensitive, at least, to respond-

ents perceptions of their parents'economic standing. Only

3 percent said that they were not able to answer the question.

This approximate distribution placed 67 percent of the re-

spondents in a family of middle income or higher. This

- 39 -

picture fits well with the educational attainment of fathers

previously reported. The rankings of fathers' incomes did

not differ substantially by the field of the respondents,

although the social scientists came from slightly higher in-

come ranks than did the natural sciences. (See Table III-5.)

(in percent )

Psychology Sociology Biology Chemistry

Less than High
School Teacher 15 22 24 21

Same as High
School Teacher 16 14 18 25

More than High
School Teacher 54 57 42 42

Don't Know 16 7 16 13

(N) (115) (125) (141) (79)

A third indicator of socio-economic status of families

of orientation of the respondents was the occupation of the

father. Occupation is, perhaps, the most commonly used in-

dicator of stratification ranking in sociology. Table III-6

compares the rankings of our respondents' fathers with those

for heads of husband-wife households between the ages of

- 40 -

35 and 44 for the 1940 and 1950 censuses. Since the specific

content of the categories has altered somewhat from census

to census, and since we coded responses to the father's oc-

cupation question on the 1970 census categories, the compar-

ison is only approximate. For full data on the distribution

of the respondents' fathers by occupation, see the Appendix.

(in percent )

4 441 .
0 r 00 00
m0 O mo
r)a) 0 0)

Professional, Technical and
Kindred Workers (includes
executives and semi-profes-
sionals) 38 6 9

Managers, Officials and
Proprietors except farm 30 13 14

Clerical, Sales and Kindred
Workers 11 13 12

Craftsmen, foremen 15 18 22

Operatives and Kindred 4 19 21

Laborers 1 24 19

Service Workers 1 6 4

- 41 -

The census data in Table III-6 are from American Families

(Glick,1957:95). It is clear that our respondents' fathers,

with especially heavy representation in the professional and

business categories, had higher ranking occupations than

their cohorts in the general population. From the previous

analysis of educational attainment, income level, and occu-

pational status, it can be seen that our respondents parents

attained uniformly high rankings on the three indicators.

The majority of our respondents came from middle class fam-


Maternal Role Models

A key element in several of our hypotheses concerning

female achievement was the special effect a mother can have

on a daughter's motivation and success in achieving. Several

questions in the instrument were designed to reveal if the

mothers of the women in our sample were positive role models

for female achievement. We shall now discuss the general

findings concerning education, occupation, and community in-

volvement of the mothers of our respondents.

The clearest model of a diversified role for adult mar-

ried women is available to a young girl whose mother works

while she is growing up. We therefore asked if our respond-

ents mothers had worked outside the home after marriage, and

- 42 -

if they had worked outside the home while the respondent was

growing up. Table III-7 shows the percentage distributions

on these two questions. The percent of mothers who never

worked was excluded from the table.


Worked After Worked While Respondents
Marriage Were Growing Up

Worked Part Time 20 18

Worked Full Time 38 28

(N) (478) (477)

According to Glick in American Families (1957:90-91),

In 1940, one out of every eight wives of family
heads was in the labor force but by 1950 this
ratio had increased to one out of every five ...
According to 1950 data, the labor force partici-
pation rate reached a low point of about 20 per-
cent during the time when wives were most likely
to have small children at home, then rose to one-
fourth for married women about 40 years old.

If we analyze the percent of mothers working after

having children for our respondents by sub-field (Table III-

8), we find that the mothers of the social scientists were

more likely to have worked, though all of the sub-fields

showed a high proportion of mothers who had worked while

having children. The percent of mothers who never worked

was excluded from the table.

- 43 -


Psychology Sociology Biology Chemistry

Worked Part Time 16 18 23 11

Worked Full Time 42 29 23 35

(N) (121) (137) (142) (78)

It is easy to see from Tables III-7 and III-8 that many of

our respondents had mothers who not only had had the exper-

ience of working after marriage, but who acted out this dual

role while their daughters were growing up and forming their

own definitions of appropriate sex-role behaviors. Astin

(1969:25) stated that for her sample one-fourth of all the

women reported that their mothers worked while they were

growing up. Astin considered this a high degree of career

orientation. Her respondents all received their doctorates

in 1957 and 1958 (though they varied in age). Our sample

reflected an even higher degree of career orientation of

the mothers of respondents, with almost half (46 percent)

working at least part time while their daughters were grow-

ing up.

- 44 -

The very fact of a mother working in a white, predomin-

antly middle class home is a significant influence on her

daughters. We should also consider the type of work she was

doing. Astin (1969:25) reported that 60 percent of the

mothers who worked while their daughters were growing up

were engaged in professional and managerial occupations.

For our sample, the breakdown differed. Approximately 47

percent of the mothers worked at jobs in the top four ranks

of jobs (professionals, executives, semi-professionals, ma-

nagerial and proprieterial), but most of these were clerical

workers (25 percent) or involved semi-professional jobs often

stereotyped as "women's work." (See Appendix for complete

distribution.) We can make an approximate comparison between

the type of occupation the mothers of our respondents were

working at and the distribution of females in the labor force

14 years old and over in 1940 classified into social-economic

groups (U.S. Bureau of the Census,1940:187). Clearly the

mothers of our respondents were usually working at jobs

higher in status than the jobs of their contemporaries in

the labor force. The distribution of mothers' jobs by status

did not vary significantly by the later choice of social

versus natural science career for the daughters.

- 45 -


Mothers of
OCCUPATIONS* Respondents 1940 Census

Professional Persons (includes
semi-professionals and executives) 39 12

Proprietors, Managers, and
Officials 9 7

Clerks and Kindred Workers
(Sales) 35 28

Skilled Workers and Foremen
(Craftsmen) 4 1

Semi-skilled Workers
(Operatives) 7 28

Unskilled Workers (Laborers
and Service) 6 25

*Other categories of ours which are collapsed to fit 1940
specifications appear in parentheses next to 1940 social-
economic categories.

The mothers of our respondents not only worked more

often, and at better jobs than their contemporaries, but

they also had attained higher levels of education than their

age cohorts in the general United States population. Accord-

ing to the 1950 census (U.S. Bureau of the Census,1950:Tables

5,8), the median years of school completed for white, married

women living with their spouses between the ages of 45 and

- 46 -

54 was 9 years. The median years of education for the mothers

of our respondents is 11 years. For the full distribution

of educational attainment of the mothers of our respondents

see.the Appendix. Also of note is the fact that 27 percent

of the mothers had higher educational attainment than their

husbands, and 59 percent had at least an equal education to

that of their husbands. Almost half (48 percent) had con-

tinued their education beyond the high school diploma, and

one-fourth had completed a bachelors degree.

There have been attempts to link daughters different

choice of life style to extra-familial activities partici-

pated in by their mothers. For example, it has been sug-

gested that mothers who do not work outside the home, but

who participate in voluntary community organizations in an

active manner, encourage their daughters to be volunteers

rather than career women. In our questionnaire, we asked

about the community involvement of mothers in the following

manner: "Was your mother ever involved in community acti-

vities as an active participant or leader (church groups,

welfare rights, political action)?" Table III-10 shows the

community involvement of mothers by sub-field of the daugh-

ters. Although there are more active members from sociology

than any other group, there seems little relationship between

- 47 -

community involvement for mothers and entrance into the na-

tural rather than the social sciences by daughters.

(in percent )


Not involved
A member only
Active member
Mother a leader

Psychology Sociology Biology Chemistry

37 37 38 40
22 17 20 21
28 23 28 23
13 25 14 16
(120) (133) (142) (80)

What does emerge, however, is the mothers of our respondents

were participants in community activities. But are the

'joiners' different women from the 'careerists?' Table III-

11 shows the results of cross-tabulating working mothers

with those who were active in community activities. If the


MOTHER WORKED: Mother Not Member Mother Mother
Involved Only Active Leader N

Never 34 24 25 18 (252)
Part Time 38 20 24 18 ( 87)
Full Time 44 12 28 16 (124)

- 48 -

mothers who were involved in volunteer work were also the

employed, then it would seem that the two models were not

mutually exclusive. Moreover, it may be that any enrichment

of the female role model leads to higher achievement on the

part of daughters. Thus, any maternal activity outside the

home would encourage consideration of broad options for life

styles of daughters. Table III-ll shows that 42 percent of

those who worked part time and 44 percent of those who worked

full time were either active members or leaders in community

activities. In addition, 34 percent of those who had never

worked after having children also had no memberships. Thus,

although those who were working full time were also the most

likely to be uninvolved in community activities (44 percent),

it is clear that working and community action often go to-

gether. These findings tend to refute the idea that volun-

teerism and career commitment are mutually exclusive. The

mothers of our respondents emerged as active and energetic

women who provided models of community involvement and work

experience for their daughters. They must have been very

well organized and committed women. This is especially ap-

parent when we recall that many of them displayed this com-

petence in several worlds when it was not fashionable to do

so--between 1930 and 1950.

- 49 -

At the close of the questionnaire a space was left for

comments on the significant events or persons who influenced

the respondents decision to obtain the Ph.D. and to pursue

an academic career. Of the 79 percent who chose to comment,

131 stated that their mothers had encouraged them, or that

they wished to be like a mother who worked. Another 12

women said that their mothers discouraged them, or that they

rebelled against their mothers role as housewife. In sum,

our respondents reported behavioral and ideological encour-

agement by their mothers in the pursuit of their own careers.

An important demographic feature of the sample, which

will be discussed in greater detail in a later chapter, is

birth order. Most birth order studies have concluded that

first-born and only children generally achieve more than

later siblings (Walters and Stinnett,1971:123). For our

total sample, 54 percent were only or first-born children

(264), 27 percent were second-born (129), 10 percent were

third-born (49) and 15 percent were born fourth or later.

Astin (1969:26) reports that:

These proportions were similar to men doctorates
(Bayer,1967). The actual proportion of women
doctorates who were first-born and those who were
born fourth or later was somewhat greater than
can be accounted for by chance, a finding that
can probably be interpreted in economic terms:
that is, the first and the last child usually
benefit because greater financial resources are

- 50 -

available to the family at the time these child-
ren enter college (Bayer,1967).

We tabulated the number of our respondents who were the

youngest in their families. For our respondents, 220 or

45 percent were the youngest in their families. If we sub-

tract from this the 18 percent who were also only children,

this leaves 27 percent who were youngest among siblings.

Until now, we have discussed our respondents in their

families of orientation. We shall now consider their own

personal attainments, and their families of procreation.

Of concern is their academic achievement, their marital

status, the socio-economic status of their husbands, and

their children.

Respondents Academic Achievement

Two measures of achievement were included in the ques-

tionnaire. One was a ranking of the graduate faculties in

the departments where our respondents received their Ph.D.

degrees. The second was a tabulation of the ranks they had

achieved at their age. Table III-12 shows the rankings of

graduate faculties by sub-field for the universities of which

the respondents are alumnae. For a discussion of the rank-

ing system, see Chapter II. In the case of each field, two-

thirds of the respondents attended a school with top ranking.

- 51 -

ATTAINED THE PH.D. (in percent )

Psychology Sociology Biology Chemistry
Excellent 75 63 69 70
Very Good 11 7 14 16
Good 5 14 10 10
Not Ranked 10 16 7 5
(N) (124) (138) (143) (80)

In terms of professorial rank attained, our women have been

promoted as they grew older. Table III-13 gives the distri-

bution of our respondents by present academic rank control-

ling for the age of the respondent.

(in percent )

AGE Assistant Professor Associate Professor N

Under 25 90 10 0 ( 23)
26-35 51 41 8 (268)
36-45 10 37 53 (127)
46-55 10 34 56 ( 46)
56-65 0 12 88 ( 16)
66-70 0 0 100 ( 4)

- 52 -

Families of Procreation

In addition to their academic responsibilities, over

half of our respondents are currently married (55 percent),

and an additional 14 percent have been married and are cur-

rently divorced, separated, or widowed. However, almost

one third of the women (32 percent) reported that they had

never been married. Thus, many more of them than their co-

horts in the general population have remained single. Ac-

cording to the 1960 census, in the age cohort of 40-44 years

86 percent of all women were married (Astin,1969:26). Mar-

riage and divorce rates varied considerably in the sample

by field of specialization, with the natural scientists hav-

ing a lower rate of marriage than the social scientists.

Table III-14 shows this comparison.


STATUS Psychology Sociology Chemistry Biology

Never Married 30 20 36 44
Currently Married 56 63 54 48
Divorced 7 10 4 7
Separated 2 4 1 1
Widowed 6 4 5 1
(N) (124) (137) (80) (142)

- 53 -

Epstein (1970:95) reports that working women in general are

far more likely to marry today than in the past. However,

the women at the top of their professions show a higher in-

cidence of being single.

A far higher percentage of the unmarried, compared
to men, tend to be found in the ranks of the pro-
fessions. Even in 1960, those women employed in
the scientific and engineering fields were consid-
erably less likely to be married than men--two out
of five women scientists as contrasted with four
out of five men.

In Astin's sample (1969:27), 32.5 percent of those in the

natural sciences had never married, while 38 percent of

those in the social sciences were single. This is the re-

verse of the relationship noted by Epstein and found in our

sample, that those in more 'masculine' occupations have

lower marriage rates. Despite these variations, the gen-

eral finding is the same--women in the professions marry

less often than other women in our society, though the trend

is toward more frequent marriage for them. Of the women in

our sample who had been married at least once, 84 percent

had had one marriage, 15 percent had had two marriages, and

less than 1 percent had had more than two marriages.

By all the socio-economic measures of status included

in the questionnaire, the husbands of our respondents had

high ratings. Ninety-nine percent of the husbands had com-

- 54 -

pleted some college and 85 percent had completed a graduate

degree. These high rankings did not differ by the field of

the wives. While it is true that all of the wives had gra-

duate degrees, the educational level of these men is still

extraordinary. Men could have traded financial success for

educational attainment and still have had as high a status

as their wives. In the case of our respondents the husbands

shared an interest in academia as well as an equivalent so-

cial status. The distribution of husband's occupations (see

Appendix) confirmed that very few of them were in business.

Most were professionals. In Table III-15 we tabulated the

number of Ph.D.'s and medical doctors among the husbands by

field of the wife. We also found that 184 husbands were

professors and an additional 25 were working in research.


Psychology Sociology Chemistry Biology
Ph.D. 47 60 34 45
M.D. (academic
medicine) 7 7 4 1 11

(N)* (88) (110) (81) (51)

*N's refer to total number of husbands in each field.

- 55 -

The table clearly demonstrates that most of the husbands are

professionals, who are involved in the academic world to-

gether with their wives. Astin (1969) suggested that such

marriages to others in academia are encouraged by marriage

during graduate school, when potential mates are found in

the university community.

Of the 324 women who reported that they had ever been

married, 240 had also become mothers. In their families,

as in those of their parents, the two-child family predomin-

ated, though 32 percent of those who had children had three

or more. Small families were the rule for all four sub-fields.

It is important to note that 245 women, or 51 percent of our

respondents, were childless. Even after we have taken into

consideration the fact that 154 of these have never married,

one third of the married women were childless. No doubt

many were still of child-bearing age and would have children

in the future. Still we have confirmed Astin's findings

(1969:29) that woman doctorates have lower fertility rates

that those of their contemporaries in the general population.

She found that the proportion of married women doctorates

who were childless was 28 percent (compared to our 30 percent)

and noted that this was "twice as large as the proportion

of childless 40-44 year old women in the general population"


- 56 -

We were interested in finding out if the women felt

that their husbands were a positive influence on their at-

taining the doctorates. Although we had no direct questions

on this subject included in the questionnaire, there are

two indicators of this influence. We did ask if the hus-

band's mother ever worked as a general measure of exposure

of the husband, before his marriage, to divergent women's

roles. The actual question specified work outside the home

after having children so as to directly tap experience re-

lating to conflicting women's roles. We found that of the

317 women who answered the question, 54 percent reported

that their husband's mother never worked, 20 percent that

she had worked part time and 26 percent that she had worked

full time after having children. Thus, as in the case of

the women themselves, many respondents' husbands had child-

hood experience with mothers who provided models combining

career and home management. A second indicator of husband's

role in achievement emerged from the coding of the comments

at the close of the questionnaire. Before discussing these

comments, we state two qualifications. First, our major

focus was on structural factors in the family of orientation

which led to achievement and not on later influences. Second,

many of our women were unmarried or had completed their

- 57 -

degrees before marrying. In these cases there either could

not have been any influence from husband, or the influence

would have been limited to career achievement and advance-

ment after attaining the Ph.D. Despite the above limita-

tions, of the 375 women who chose to comment on significant

factors in their decision to obtain the Ph.D. degree and to

pursue an academic career, 130 mentioned the role of people

other than their parents in the decision-making process.

Of these 130, 75 (or 58 percent) cited the husband's influ-

ence. Of these, 70 mentioned their husband's encouragement

and 5 noted their husband's discouragement. Some typical

comments were:

My husband is wonderfully supportive of my career,
thus I have not 'interrupted' it for child bearing.
With his help it has been possible to continue even
with young children.

Most important was my husband's attitude--his moral
and financial encouragement have been the most help-

After we had been married 15 years my husband wanted
to get out of management and pursue his mathematical
interests. He entered graduate school ... actually,
my husband would not give up his job and change pro-
fessions unless I would go to school, too. We were
looking for a complete change of life style which
included both of us actively involved outside the
home. Now, I would not have it any other way--but
without the initial pressure from these three men I
would have continued to think I could not obtain an
advanced degree and would have settled for a much
lower goal.

- 58 -

Throughout this chapter we have attempted to describe

our respondents in the most general fashion, and yet to por-

tray them as real women in real life situations faced by

many women. We have seen that their decisions to opt for

a minority life style in their culture went together with

certain demographic characteristics of their families of

orientation. These included disproportionately small fami-

lies, urban settings, certain ethnic and religious affilia-

tions, middle class socio-economic status, parents with

above-average educations, and fathers with higher than aver-

age occupations. Their families often included mothers who

worked after having children and who were active in voluntary

associations. Our respondents studied for their degrees in

the top ranked institutions in their fields, and have gone

on to attain high academic ranks, often while fulfilling

the roles of wife and mother. Their families of procreation

are distinguished by the high educational training received

by their husbands, their shared experience of the academic

world, and their low fertility rate. One third of the re-

spondents never married, and 50 percent were childless.

Those who had married seemed to have stable relationships

as evidenced by the low rate of divorce they reported. If

we return to the questions posed at the beginning of this

chapter, we find that they have been answered. Our respondents

- 59 -

differed from their age cohorts in the population in degree,

but not in kind. They came from stable homes and they are

building stable homes for themselves. Most have sought to

combine career, marriage, and motherhood, often in imitation

of their own mothers. We have now been 'introduced' to our

respondents and in Chapter IV we shall go on to consider

their collective story in relation to our own specific theo-

retical interests and hypotheses as they were delineated in

Chapter I.


The hypotheses to be tested in this study fall into

several categories. First there are those which relate spe-

cifically to the relationship between the sex structure of

the nuclear family and female achievement. Second, we are

concerned with family size and achievement. Third, there

are general birth order propositions which have previously

been tested on men and which we seek to apply to the case

of female achievers. Fourth, there are hypotheses which

have been considered in the previous chapter, including the

relationship between socio-economic background and achieve-

ment and that between maternal employment and female achieve-

ment. This chapter focuses on those hypotheses which relate

female achievement to the sex composition of the nuclear

family, ordinal position, and family size.

Sex Composition of the Nuclear Family

Our original hypothesis was that,in the absence of male

children, parents would consider the need for fulfillment

of the American dream of upward mobility for their children

- 60 -

- 61 -

more important than accepted sex-role definitions. Conse-

quently, they would allow, encourage, or push one of their

daughters to achieve in the occupational sphere. Should

these conjectures be correct, a sample of female achievers

should have contained a disproportionate number of women

from all-female sibling families. Table IV-1 shows the per-

centage of all-female sibling families found in our sample,

and next to that a theoretical expected percentage derived

from a probability model. The calculation of the expected

percentages was discussed in Chapter II.


Respondents Percentages

All-Female 46 44

Mixed-Sex 54 56


Our sample as a whole did not contain a much higher propor-

tion of women from all-female families of orientation than

expected, although the differences fall in the expected dir-

ection. Table IV-2 shows the number of respondents from all-

female families, by family size, and compares these percentages

- 62 -

with those expected for families of these sizes based on an

ideal probability distribution. X2 was computed separately

for each family size.
(in percent )

COMPOSITION 2 3 4 5 6 7 8+

Mixed 0 52 71 77 79 93 100 100
(92) (83) (48) (22) (13) ( 7) ( 9)
All-Female 100 48 29 23 21 7 0 0
(87) (84) (29) (14) ( 6) ( 1)
Expected %
All-Female 100 50 25 12.5 6 3 1.5 -

X2 = 4.5, p (.05 + X2 = 8, p< .005

As family size increases, there is a greater likelihood than

expected of coming from an all-female family. For families

of four and five children, the number of respondents from

all-female families is significantly different from that ex-

pected, and in the direction hypothesized. If we look at

the sibling sex composition of the family of orientation by

field of specialization of the respondents, we find that more

of those specializing in the social sciences come from all-

female families. (See Table IV-3.) However, Table IV-4

shows that this was accounted for by the greater proportion

of small families of orientation among the social scientists.

- 63 -

(in percent )




Psychology Sociology Chemistry Biology

52 52 60 56

48 48 40 44
(124) (138) (80) (143)



Psychology 21 39 21

Sociology 17 38 2C

Chemistry 14 36 2E

Biology 18 32 2C

3 4





5 6

5 3

4 2

3 4

10 3

These results suggest that choice of field of specialization

by women from all-female families was based on factors unre-

lated to the sibling sex structure of the nuclear family.

Most probably, they were channeled into certain fields by

personal interest, exciting teachers, or excellence in these

fields in high school, and other factors mentioned in their

comments. The family of orientation may have fostered a


1 (124)

3 (138)

* ( 80)

3 (143)

- 64 -

climate in which women were encouraged to go on to higher

education and career attainment, but it did not necessarily

specify participation in particular 'masculine' fields as

true proof of achievement. In fact, daughters' achievement

in the social sciences may have eased the conflict between

sex-role definitions and the need for a child who achieved,

since participation in the social sciences on a professional

level was more acceptable than in the natural sciences.

We next ask, are there any sub-groups within our sample

of achievers which showed a greater relationship between sex

structure of the nuclear family and achievement than we would

have expected by chance? To examine the above possibilities,

we controlled the relationship between sibling sex structure

of families of orientation and family size and looked at the

distribution of our respondents by religion and by ethnicity.

Table IV-5 examines the differences in sibling sex structure

of family of orientation by religious affiliation of mothers

of respondents. From Table IV-5 it appears that the Catholics

were producing less achievers from all-female families than

expected, and that the Jews were producing more. This was

partially due to the effect of family size. The role of fam-

ily ideology may also have been crucial. Of course, this

is only a suggested interpretation. Jews are known for their

- 65 -

commitment to achievement, while Catholics are more influenced

by the traditional sex-role definitions exemplified by Mary

and stressed by the Church.


Protestant Catholic Jewish Other Agnostic
Mixed 56 60 49 60 42

All-Female 44 40 51 40 58

(N) (279) (69) (105) (10) (19)

It is also of note that several of the respondents who wrote

in "none" or "agnostic" under religion of mother also wrote

in Jewish under ethnicity. What we may have represented

here, then, is a secularization curve. The more secular the

family, the more committed to achievement, and the less com-

mitted to traditional sex-roles. Thus, the continuum from

religious to secular within religious categories goes:

Catholic, Protestant, Jewish, None or Agnostic (made up of

members of all groups mentioned previously who have been

completely secularized). Table IV-6 shows that even after

controlling for family size the percent of Jewish respondents

from all-female sibling families sized two and four is higher

- 66 -

than that for the other groups. The distinction between

Catholics and Protestants is not clear. The category of

none or agnostic was not included here because the numbers

were so small..The percentages for mixed sex sibling fam-

ilies were not shown in the table.

1 2 3 4 5 6
Catholic 100 38 40 29 66 25
(13)* (6) (4) (2) (2) ( 1)

Protestant 100 49 29 21 19 0
(48) (42) (17) ( 8) ( 3)

Jewish 100 54 33 37 25 0
(19) (28) (8) (3) ( 1)

Expected percent
per family size 100 50 25 12.5 6 3

*"N"'s in parentheses refer to number of respondents
from all-female sibling families for each religious group.

As we noted in Chapter III, one third of the respondents

reported that both their parents were members of some ethnic

group. In order to find out if ethnic affiliation affected

the relationship between sibling sex composition of the fam-

ily and female achievement, we cross-tabulated sibling sex

composition by size of family with a control for ethnicity.

The results are found in Table IV-7.

- 67 -


1 2 3 4 5
Mixed Sex

East and South

West and North


25 100





All-Female Siblings

East and South

West and North













( 7)




( 5)

Total Sample
Percent All-Female 100 49 31 23 22

Total Percent All-
Female of Those Who
Checked An Ethnicityl00 50 25 26 42








( 7)

- 68 -

TABLE IV-7 (cont.)

1 2 3 4 5

Expected Percent
All-Female 100 50 25 12.5 6

In this table, ethnic groups were combined by geographical

areas. Eastern and Southern European ethnicity includes those

who checked Polish or Italian, while Western and Northern

Europe includes those who indicated affiliation with Irish,

English, or German ethnicity. Those who were included in

the "other" category were predominantly of Northern and

Western European descent as well. Therefore, for the calcu-

lations in Table IV-7, "other" was combined with Western and

Northern Europe. Jews were separated into a third category.

The pattern which emerged in Table IV-7 was a positive link

between ethnicity and female achievement for women from all-

female sibling families whose ethnic identification was with

Western and Northern Europe, as well as for the Jews, espe-

cially in families with more than three children. In addi-

tion, it appeared that for those who checked any ethnic iden-

tification and who came from families of four or five child-

ren, there was a greater likelihood of coming from an all-

female family than for those who checked no ethnic affilia-


- 69 -

Blau and Duncan (1967:240) in their study of male

achievement in the American occupational structure found

that members of white ethnic minorities from Western and

Northern Europe fared better in occupational success than

the dominant majority. This was not true for Poles or

Italians. They attributed the achievement of the Western

and Northern Europeans to the extra achievement motivation

felt by sons of immigrants, and the lower rates of achieve-

ment by Eastern and Southern European immigrants' sons to

discrimination. Their finding seemed to parallel our own

concerning the push for achievement on the daughters of

Western and Northern European immigrants who still felt

ethnic ties. Padan-Eisenstark (1972) found that career women

in Israel do not come disproportionately from all-female sib-

ling families. Using similar logic to our own, she hypothe-

sized that women from all-female sibling families would be

disproportionately represented in a sample of career women

in Israel. Her findings directly contradict our own. The

Jews were the one group we consistently found to come dispro-

portionately from all-female sibling families, and all of

Padan-Eisenstark's sample are Jewish.

When we combined the Blau and Duncan findings on sons

of immigrants from Northern and Western Europe with our own

- 70 -

findings on ethnic groups and the findings of the Israeli

study, a contextual explanation emerged. Members of minority

groups often have felt under pressure to achieve. Jews in

America, and children of immigrants who came from a culture

where they were the majority to one where they were a minor-

ity, have felt that pressure. Jews in Israel, on the other

hand, are members of a majority group whose children are

expected to fill all of the occupational slots in the country

and develop no such minority pressure for achievement. But

how can we explain the lack of similar pattern among those

identified with Southern and Eastern European ethnic groups?

If more of the members of these groups were from Catholic

families, then the ideological pull of the traditional sex-

role definitions may have out-weighed the need for achieve-

ment. If we conceive of female achievement as the end-product

of the resolution of conflict, then it is the relative weights

of the countervailing forces which lead to the emphasis on

sex-role definitions or to the dream of upward mobility.

For the family which placed greater emphasis on sex-role de-

finitions than on achievement, upward mobility may be fore-

gone or mobility may be 'achieved' through marriage by a


- 71 -

In our findings on sibling sex composition of families

of female achievers we noted that the groups with the lowest

birth rates (those of Jewish and white Protestant background)

had.the highest proportion of achievers from all-female sib-

ling families even though the improbably large proportion

of female achievers came from large all-female sibling fami-

lies. In the absence of early sons in a family of a group

in which two or at most three children is the norm, a deci-

sion may be made to have more children in the hope of pro-

ducing a male. This behavior would indicate a strong need

for a son as potential achiever. This strategy failing, the

parents put pressure on one of the girls to achieve. In

other words, a large family size in a white Protestant mid-

dle class home or in a Jewish middle class home coupled with

the absence of male children indicated or coincided with

the parents' strong need for an achieving child and consequent

stress on their female children to achieve. The two pre-

ceeding discussions were suggested by our findings, although

not conclusively supported by them. They both pointed to

interesting future lines of research.

- 72 -

Family Size

According to Blau and Duncan (1967:298), men from small

families (less than four children) achieve more than those

from larger families. Turner (1962:122) found that boys

from large families "are less likely than those from small

families to have high ambitions, regardless of background."

Turner also noted that "the insignificant coefficient of

-.05 for women fails to support the same conclusion, however.

Why the relationship should not apply to women students as

well as to men our data do not clarify"(1962:122). We have

Already shown in Chapter III that our achieving women came

disproportionately from small families. However, let us

look at achievement within the sample. An indicator of

achievement included within the instrument was professorial

rank attained. When we controlled present academic rank by

age and examined family size by rank, we found that women

from smaller families attained higher rank only at the cul-

mination of their careers. Tables IV-8, IV-9, and IV-10

illustrate this clearly. The first two tables (IV-8, IV-9)

seem to substantiate Turner's finding while the third (IV-10)

follows the Blau and Duncan finding.

- 73 -

(in percent )

FAMILY SIZE Assistant Associate and
Professor Full Professor (N)

1-4 Children 54 46 (261)

5+ Children 61 39 ( 28)

X2 = .457 (not significant)

(in percent )

FAMILY SIZE Assistant and
Associate Professor Full Professor (N)

1-4 Children 49 51 (106)

5+ Children 52 48 ( 27)

X2 = .159 (not significant)

(in percent )

FAMILY SIZE Assistant and
Associate Professor Full Professor (N)

1-4 Children 28 72 ( 57)

5+ Children 63 37 ( 8)

X2 = 3.7, p <.05 (at 3.84)

- 74 -

A second hypothesis advanced by Blau and Duncan (1967:

300) was that only children achieve even more than children

from small families regardless of the birth order of the

latter. When we looked at present academic rank by age by

family size for the only children among the respondents, we

found that there was no greater achievement for only child-

ren than for respondents with siblings.


(Assistant Professors)
Only Children With Siblings
23-35 95 90
36+ 5 10
(N) (36) (140)

X2 is not significant


(Associate and Full Professors)
Only Children With Siblings
23-35 37 43
36+ 63 57
(N) (61) (257)

X2 not significant

- 75 -

Tables IV-11 and IV-12 showed that women who were only child-

ren did not attain significantly higher ranks than their age

cohorts with siblings.

A third hypothesis advanced by Blau and Duncan (1967:

300) is that men from small families started higher and ad-

vanced further in their careers than those from large fami-

lies. Did this hold for women as well? Our indicator of

"starting higher" was the school ranking of the institution

from which the woman received her Ph.D. Table IV-13 shows

the results of the cross-tabulation of number of siblings

with school rank.


Excellent Very Good Good Not Ranked (N)
1-4 Children 70 12 7 12 (427)

5+ Children 69 10 10 12 ( 52)

X2 = .641 (not significant)

For our sample, women from small families did not start out

at higher ranked schools than those from larger families.

We have already shown that women from smaller families did

eventually attain higher rank than those from larger families,

- 76 -

parallelling the second part of the Blau and Duncan findings.

In sum, we have seen that family size did affect achieve-

ment for our women. First of all, most of the respondents

came from families of smaller size than their cohorts in the

population. Second, those from smaller families eventually

attained higher rank than those from larger families. When

the only children were considered separately, they followed

the small family trend. No advantage was found for any par-

ticular family size with regard to rank of school attended.

Perhaps this is a reflection on the superior quality of the

total sample.

Birth Order and Achievement

Most of the research linking ordinal position and

achievement has tested male subjects. Sometimes, the find-

ings for males have been generalized to both sexes. (See

Chapter I.) In our theoretical argument, it was agreed that

birth order would affect achievement; however, the automatic

translation of the results for men into those for women was

avoided. In fact, we suggested that birth order might have

a different meaning for men than for women. We suggested

that a large percentage of youngest daughters might achieve,

rather than those who were eldest. We also hypothesized that

birth order may have had different meanings within different

- 77 -

contexts--specifically within the all-female versus the

mixed-sex family. Other hypotheses to be tested related to

the advantages of coming from the extreme positions in the

family, rather than being a middle child. The remainder of

this chapter will be concerned with the effect of ordinal

position on females as reflected by our sample of female


The general findings on birth order for the total sample

were mentioned in Chapter III. There we noted a large pro-

portion of first-born and only children, as well as a repli-

cation of Astin's (1969) finding that there was an unusually

high percentage of children born fourth or later, and that

27 percent were the youngest of their families after the

elimination of the only children from the tabulations.

Table IV-14 shows the disproportionate number of eldest child-

ren in the sample after elimination of the only children.

For this table, the expected frequencies within family size

were calculated using a probability model. Thus, the pro-

bability of being a middle child in a three-child family was

one third. Two-child families were not included in the table.

Within the two-child families, 54 percent of the respondents

were oldest and 46 percent youngest.

- 78 -


3 4 5

Oldest 45

Middle 27

Youngest 27

(N) (102)

X2 = 19.52, p (.001










= 4)

Let us now consider if birth order affected girls dif-

ferentially within the context of different sex structures

of the nuclear family. Did the eldest have a greater advan-

tage if she came from a mixed-sex or an all-female sibling


(in percent )


Mixed-sex 42 58 (261)

All-Female 50 50 (137)

- 79 -

It appears from Table IV-15 that it was a greater advantage

to be an eldest daughter in an all-female sibling family

than in a mixed-sex sibling family. However, this could be

accounted for by remembering that the all-female sibling

families were smaller than the mixed-sex sibling families,

thus giving a greater a priori probability of not being

eldest. The expected relationship between being first-born

and achieving held for our women, and sex structure of the

sibling group did not affect it.

We looked at the comparative percentages of youngest

children within the context of all-female and mixed-sex

families of orientation for our sample.

(in percent )

Mixed-Sex All-Female
Not youngest 64 63
Youngest 36 37
(N) (264) (134)

Table IV-16 shows no difference in placement origin for the

respondents on the basis of sex structure of the nuclear

family. Blau and Duncan (1967:307) suggested that children

- 80 -

in extreme positions are the most successful. Moreover,

they stated that this relationship held regardless of socio-

economic status, family size, or the existence of an older

brother. For our total sample, we had the following distri-

bution by position: 18 percent only children; 36 percent

eldest (excluding only children); 18 percent middle children;

and 27 percent youngest children (excluding only children).

Of course, for the 36 percent who came from two-child fami-

lies, there was no possibility of being a middle child.

Using school rank as an indicator of achievement, we

tabulated the distribution of ordinal position of the respond-

ents among the school ranks.

(in percent )

POExcellent Very Good Good Not Ranked (N)

First-born 67.4 14.8 6.4 11.4 (264)
Second-born 72.0 5.4 13.2 9.3 (129)
Third-born 77.6 6.1 12.2 4.1 ( 49)
Fourth-born 55.6 7.4 22.2 14.8 ( 27)
Fifth-born 40.0 60.0 0 0 ( 5)
Sixth-born 82.0 0 9.0 9.0 ( 11)

From Table IV-17, it did not appear that oldest children or

only children had higher achievement than the middle child-

ren in our sample. Perhaps this was due to a screening

- 81 -

effect discussed by Blau and Duncan (1967:308-9). Male

middle children who had made it past the bachelor degree

were apt to be more successful in graduate school than those

from smaller families. In effect, then, the B.A. was the

biggest hurdle for a middle child to accomplish. Only the

truly superior ones made it, since they often had less emo-

tional or financial support from their families than first-

borns did. If this applied to women who chose to achieve

in 'masculine' terms, then our women, who were all past the

doctoral level, were reflecting this phenomenon. There

were less middle children than eldest or youngest in the

sample because less female middle children made it past the

B.A. than those in other positions. However, those that made

it past the bachelor degree went to schools of at least as

high a rank as any other group in the sample. This same

selection pattern is mirrored by the high ranks of institu-

tions where our respondents received their Ph.D.'s compared

to male doctorates. Fewer women than men made it past the

B.A., but those who did often excelled in graduate school

(see Chapter III).

Insum, among our sample of female achievers there were

a high proportion of first-born and only children, as well

as of youngest children. However, ordinal position did not

82 -

have different meanings for women from families of different

sibling types. In fact, it had no meaning in terms of

achievement within the sample. Middle children were less

likely to make it through graduate school, but those who did

attain the Ph.D. achieved at least as well as their peers

who were first-born or youngest children.


Through this study of female achievement, we set out

to explore a part of the socialization process which gives

impetus to a woman to achieve through a career instead of

or in addition to marriage. Our specific focus was the

effect of the sex composition of the sibling group of the

nuclear family of the respondents and the consequences of

the make-up of that sibling group for the achievement goals

of the parents. The 'push' toward achievement was assumed

to come originally and most powerfully from the parents.

Our analysis of the data, then, was really an analysis of

the countervailing forces at work on the parents and dir-

ecting the parents' effect on the daughters. Those forces

influencing the parents which emerged during the study were

extent of ideological commitment to traditional sex-role

definitions and belonging to a minority or immigrant group

which made the need for achieving the American dream more


Our more specific findings included the following.

First, socio-economic status of the family of orientation

- 83 -

- 84 -

was related to achievement. The respondents came overwhelm-

ingly from middle class families. Their parents were highly

educated, middle income couples. Their fathers had presti-

gious occupations. Second, the women in our sample very

often had diversified models of the female role. Many of

their mothers worked after they had children--and they

worked out of choice at occupations of higher status than

their age cohorts. Not only did their mothers work for

salary, but also as volunteers in community organizations.

Often they were active members or leaders of these community

groups while having full-time jobs. Third, our respondents

came from urban environments. Fourth, at least one third

of them were consciously part of ethnic groups. Fifth, they

came from stable homes where both parents were present while

they were growing up. More than 95 percent of those who

mentioned relationships with parents and the parental influ-

ence on their pursuit of an academic career in their comments

at the end of the questionnaire felt that their parents were

a positive influence on their decision.

The analysis of the hypotheses concerning the relation-

ships between sex structure of the sibling group in the nu-

clear family of the respondents, family size, and ordinal

position, and female achievement yielded the following results.

- 85 -

For the sample as a whole, there were no significant differ-

ences in sex composition of sibling groups for female

achievers. However, when the respondents were compared by

ethnicity of parents and by religious groupings, dispropor-

tionate percentages of all-female sibling families emerged

for certain groups. Thus, among Jews there were a dispropor-

tionate number of all-female sibling families for three and

four child families. For those who were affiliated with

ethnic groups, those who came from Western and Northern Eur-

ope came disproportionately from all-female sibling families

for families of size three, four and five. The finding pre-

viously noted for "religious" Jews also held true for "ethnic"

Jews. We have already discussed the possible linkage of the

findings for Jews and for children of immigrants from North-

ern and Western Europe in Chapter IV. We speculated that

when couples from groups with low fertility rates continue

to have children in order to have a son, and don't succeed,

this leads to a push toward achievement for one of the

daughters. We found that,as in the case of men, family size

had an effect on achievement. Our women came disproportion-

ately from small families, compared with their age cohorts

in the census. While no advantage was found for any parti-

cular family size in terms of rank of school attended for

the doctoral degree, eventual achievement of higher profes-

- 86 -

sional rank was related to small family size. This was un-

like the finding for birth order where those who were in

the extreme positions were most likely to have attained a

doctoral degree, but, once possessing the degree, their at-

tainment within their fields was unrelated to ordinal posi-


Throughout the analysis, comparisons within the sample

were made by sub-field. While these comparisons were inter-

esting, especially with regard to the demographic data, the

primary finding resulting from their use was that choice of

field was not related to structural factors in the nuclear

family, but to other, later forces in the respondents' lives.

Finally, there emerged from our data a composite portrait

of the life style of our respondents today. The two thirds

that were married most often had only one marriage. Their

husbands were as educated as they and shared their interest

and involvement in the academic community. They have mar-

ried later and had fewer children than their age cohorts

in the census, but they seem to be successfully combining

family life, motherhood, and academic careers. Their hus-

bands most often came from homes where their mothers worked

while raising children. Together, these couples are pro-

viding the same kind of diversified female sex-role models

- 87 -

for their children of both sexes that their mothers often

provided for them.

Future research on the effect of variant sibling sex

compositions of nuclear families can be broadened in several

ways. First, in order to perform a more direct test of the

links between sibling sex structure of the nuclear family

parental ideology, and female achievement, it would be

useful to do a study of parent-daughter combinations. Such

a study would include actual measures of ideological com-

mitment to traditional definitions of sex-role by the

parents. A second facit of a direct test of our hypothesis

would be the introduction of a control group of non-

achieving women, rather than the use of the census cohort

of the respondents as a pseudo control group. A third way

of enriching future studies of female achievement is to

study women who have been successful in areas other than

academia. Such women would include business executives,

artists, professionals, and politicians. One cannot gen-

eralize results of a study of academic women to success-

ful women in all fields of endeavor. Fourth, the meaning

of identification with an ethnic group, a minority group,

or an immigrant group should be explored further. Finally,

a parallel study should be carried out of all-male sibling

- 88 -

families to see if parents will "permit" one son to deviate

from accepted "masculine" occupations if there is another

brother to be the achiever. In other words, are a dispro-

portionate number of male poets, actors, elementary school

teachers, dancers, and the like from all-male sibling


We shall close this summary and conclusions with

thoughts from our respondents which illuminate some of

the findings of this study. They do not speak in per-

centages, but their message is loud and clear.

On Female Role Models:

The typical western or midwestern atmosphere in which
I grew up is much more conducive to female education
and achievement. Both sides of my family were in Ohio
or Michigan by about 1850. Both my grandmothers
graduated from small Ohio or Michigan colleges about
1900. My aunts were taking master's degrees in the
1920's. I have thought about these matters quite a
bit. I have lived 20 years in New Jersey and have
a 12 year old daughter. (from a chemist)

The "liberation" in our family occurred in my mother's
generation. She came out of non-English speaking
ghetto background and somehow became a professional.
I was proud of my mother's accomplishments and always
knew I would use my capacities to their maximum. I
was supported and pushed (sometimes too vigorously)
by my mother and her friends. Both my daughters are
in college now in pre-professional courses. Interest-
ingly, my younger sister had exactly the opposite
reaction. (from a biologist)